Whats not being said about gYFI being “non transferrable” is those people saying it dont realize that it actually is even if you write transfers out the code. What happens when i empty my account and sell the seed words through a company that does escrow to sell my veCRV or gYFI?
Edit: youre just making another token to buy and sell. You idealists can say its valueless all you want. Leave YFI as is! Big no on gYFI from me
Making it non-transferrable at least adds a fair amount of friction to sell it. This will probably always be a game of cat and mouse, but as long as it is not rampant then it probably isn’t a huge problem.
See these threads from the dxDAO about trying to prevent people from selling accounts with non-transferable REP.
That was my initial reaction too @Joey. But I assume you have some concerns that a bad actor interested in destroying the Yearn system could easily borrow YFI on a lending market, prevail on a destroying vote, return the borrowed YFI, and sit back and enjoy the carnage those actions created?
If you do not want to separate governance from YFI, which seems obvious from your comment, do you have any suggestions to reduce this risk from a bad actor? This is somewhat different than vote selling, but you may not see a plan that relies on increasing voting power over time to be sufficient. I am curious as to your ideas. Or maybe you think these risks are exaggerated.
Why not just implement similar restrictions that RenVM has? 1 epoch to enter, 2 epochs to exit, ren token controls everything (or will), and add a orangecore. Did anyone here read the new a16 article? Why are we trying to find the one true decentralized system through gYFI instead of just implementing semi-centralization such as the orangecore idea where 13 multisig holders also sign. I dont like “leaders” any more than the next but as long as they dont have a vote on their own seat Im okay with it
Yes why re-invent the wheel and risk it fail
I am familiar with RenVM, but reviewed it in more detail and looked at the a16z article on staged defentralization. I also looked at some RenVM discussion involving the change to their epochs and a delay in the relase of the bond after a node deregisters. All of it was interesting and should be kept in mind as Yearn tries to move towards greater decentralization, but it is not obvious to me how these issues about moving towards decentralized nodes are directly related to concerns that Yearn’s governing token could be borrowed and used for an evil vote that destroys the system.
Perhaps there is a solution to this problem from the RenVM plan and a staged move to greater decentralization, and if so perhaps you or someone else will help draw those connections. Until then, there are likely concerns in the Yearn community that too little YFI is staked in governance and that it could be relatively easy to borrow enough to cause harm.
Ok so let me present a scenario. Forget Ren for clarity sake.
Flash loan vote happens maliciously. We need prevention. Step 1. Require tokens to be locked for 1 month before voting.
Okay well what if they just own their tokens? Step 2. Allow yvault to vote after all other votes are done to confirm the vote so no one comes in at the last second. Also do not require yvault tokens to lock.
And finally what if the person actually pulls it off somehow? Now they want to dump YFI too before the crash. Step 3. Tokens stay locked for 1 month after voting
Edit and if we need more protection then forget YVAULT voting and just explicitly say Andre must confirm each vote unless unless otherwise unable
Edit 2 or whomever we vote CEO or multisig final decider. Again im not a fan of having leaders but its better than gYFI token split
A split of the functions would imply that governance (gYFI holders) would cede some of their rights to control use of fee income. Otherwise gYFI holders could first sell off their YFI and then pull the rug by voting to either keep revenues or spend all of them.
Thank you for connecting this to your staged decentalization comments from before. I think you are saying that we do not need to be so decentralized so that a vote is automatically implemented without some additional check or veto. I suppose in practice that is where we are – implementation requires the actions of individuals, including Andre and the multsig. And presumably they would step in and refuse to implement a harmful vote. I suppose they similarly would take action if a bad actor maliciously voted against something important. As long as these leaders are around to protect the sysrem for us (and are willing to do so), we should be ok. I know you mentioned disliking reliance on leaders, so working towards another solution would be ideal.
Only because you mentioned them, let me address some of the other points.
Locked for 1 month before voting: this may provide a disincentive to participate in governance for others and then require less voting power for the bad actor to skew the vote.
Locked for 1 month after voting: again this may provide a disincentive for others to participate in governance as well.
Passes evil vote but needs to sell YFI before impending crash: if acquired YFI were borrowed, the crash would make it easier for the bad actor to pay back the loan (return borrowed YFI). Or wait for the one month lock and return borrowed YFI. If the price of YFI crashed, the risk of liquidation would also decrease. The cost of implementing this under your hypothetical is the cost to borrow for 2 months. I don’t know if that is prohibitive. But today we only have a 3 day lock, so the borrowing cost is certainly much less.
You are correct that this suggests gYFI holders would not have the same financial rights as YFI holders. Although I am not convinced that we should, some have advocated for a gYFI token, and I was thinking about how it might work.
Imagine if gYFI holders were entitled to 10% of any amount actually passed to YFI holders. Or up to 10% if certain growth or other metrics were met. Would that give gYFI holders a reason to support the long term wellbeing of Yearn? We might think of something akin to a long-term incentive compensation program. Worried about gYFI holders changing the percentage or metrics – make it immutable or, better yet, give YFI holders limited veto rights for changes that directly benefit gYFI holders. Another option: have a board appointed/delegated by holders of each that approves governance decisions.
The idea was to prevent governance from being borrowed on AAVE, for example, and used maliciously. @Joey pointed out that Andre and others currently stand in the way of such malicious action, so perhaps for now a gYFI token is not needed if the only justification is unlikely to develop. Some pushed for a governance token to allow yield seeking uses of YFI while permitting YFI holders to participate in governance. gYFI was a potential solution for that as well.
Yes. This.
I’d rather centralize than the scenario @TheSouthSeaCompany presented happens, because I agree completely that can happen and I think it probably would. Edit: by centralize here I mean enact leaders by vote whom may also be removed by vote
YFI has to remain the sole power. It doesnt work to just keep making new power tokens. Make it harder to vote if we must such as slashing can be considered.
How about:
Each vote is given $5
Voter uses the $5 to vote
Vote either returns $10, or $5. $5 for against. $10 for for. If voter votes against the majority 20 times, now it will return $3. Voter still isnt paying, but just being rewarded less. At some point this could go negative and slashing could be as aggressive as people want.
I think there are many alternatives to making gYFI and we really need to make YFI the only token.
What do you think are the next steps? Formalize leadership roles with veto power? Increase voting power over time (with a cap)? Change the lock period after voting? Find ways to increase yield on governance staking? Expand voting to include YFI vault holders? Should we encourage or discourage yield seeking uses of YFI compared to governance uses?
Yea pretty much. Vote in 7 people to be able to jointly veto, all of whom may be voted back out with seats taking a certain amount of days to vacate.
Im against voting power increases. All the rest I’m open to.
For yield, I think it makes sense for governors to make less.
For yvault, I think it can already participate in voting if need be. Its just something to keep in mind as an added strength for us that Andre controls a vault able to vote with great power. Such as for the 2 days downtime we could have switched to a governance strategy until cream is ready
I’m not sure I understand how the YFI vault can already vote or if Andre could unwind the strategy, stake the YFI, and vote fast enough to counter a last minute malicious voter.
If governance stakers continually receive lower yield than YFI holders who do not stake in governance, we could see an ongoimg reduction of participation in governance. I am generally against that, although it is a choice every YFI holder will have to make – participate in governance or earn more yield.
Maybe a second vault could explicitly not loan the YFI. Only use it in governance, sell the yCRV, rebuy YFI automatically
It seems a little weird to have this as a separate thing from the governance staking contract.
If people want, they can already exchange their yCRV gov staking rewards for YFI.
(Barring gas costs, which can be a problem for smaller holders.)
Governance is a job. Having a yvault do it for us has 2 advantages: easy for some people and small holders, and yvault has power to sway votes to protect the network.
For investors like myself id take this option. Its extremely low risk and benefits the network im invested in
I’m still not sure I follow. How would the yYFI vault vote on proposals, and how would the stakers influence this vote? (As in, how would people in the vault signal their vote, if not via a transaction.)
100% Andre or controller multisig choice
Edit: this is a separate idea from having 7 people with veto power
Edit2: or you could state the leaning of the pool before hand and vote accordingly: anti inflation, etc.
Edit 3: theres no reason yfi couldnt have 4 yvaults each with specific strategies slightly tailored but collecting the same results
Ah, okay. So people would be essentially delegating their votes then?