Because this is being exploited right now, and I’d like to start a vote to update the premium immediately if this vote passes, which will give us plenty of time to deploy a longer-term solution.
I think this idea posted to Twitter is pretty cool: https://twitter.com/avsa/status/1480309719910731782
Reading this thread it’s clear that people are reading this as “you have to pay 100k to renew” and not “expired domains go through a Dutch auction”. Maybe the starting price should be set at 2^128 dollars, so that it’s basically at infinite.
If the start price was the Ethereum market cap, and the end price is $0.50, the decay rate is about 4% per hour (~60% per day), which is much less than I expected.
Working backwards, 50% decay per day seems a reasonable starting point; if we wanted the end price to be $5 (same as a year’s registration), the start price would be 1.3 billion dollars. It’d hit 10 million after 7 days.
Also gets rid of the “i’m priced out” argument, as everyone would be. I think that suggestion has the better story to it than upping the price to the proposed 100k.
@jalil Thanks for sharing this.
I believe this could be a working longer term solution. The current proposal is a “stop-gap”.
Though I’m in favor of the current proposal and have voted for it, my hesitation was also “how does this look to others?”. While helping out with ENS Discord, I’ve heard more than one user complain, “oh great now the name I want is expired and I need to pay $2000”. I’m thinking “wait til you see it at $100k.” (I didn’t say that though )
I understand the delegation process, that most users won’t have been following, understanding, or keeping up with ENS, so they delegate their voting power to those who resonate with how they might vote. Informed voting is important!
With that said, it looks like the proposal is going to pass. However, there’s nothing stopping us from immediately making another proposal to up the starting price to 2^128. This may have the intended effect to help people realize the starting price isn’t “ENS selling names at $100K now”. They will realize, “Oh it starts out of reach, and goes down. I’ll wait until day X of 28 days to try and buy it for a price I (or my company) thinks it’s worth.”
I agree, and that’s a good long-term plan. However, we need to fix the immediate problem first and implement a longer-term fix later.
Thinking more about this, implementing exponential decay may be easier than I thought. If we set the decay per period at 50%, we can simply right-shift the premium by 1 bit per period elapsed, and then use fixed point exponentiation to calculate the fractional part.
I suspect it’s still more than a couple of days work, though!
i agree. it will communicate to people that ENS domains are that valuable. then asking prices will shoot through the roof on listed and future listed names.
either way i agree there needs to be a sort of bot proof implementation . not good for prices, not good for onboarding.
@nick.eth do you know if there any cases of people sniping names being registered for the first time ? or if that is even possible?
If the name has never been registered before, then anybody can register it at any time. So no, sniping is not possible, or rather that term doesn’t really make sense here.
But it is possible that if you set a low gas price on your register transaction, that someone else can see that and go through the commit tx + wait 1 minute + register tx and complete all of that with higher gas price before your register tx finishes.
Since this passed over the weekend, I’m deploying the new oracle now. We will need ChainLink to whitelist it before I can start an onchain vote.
Edit: I’ve deployed the new price oracle contract at 0xa51b83e420c5f82982dc8b7f4514c9bea0b290ee. Inspection and verification welcome!
This is now live for voting onchain here
Please go vote!
Usually, in a scenario where transaction costs are zero, the most efficient outcome will occur. I got that from a law book going over control mechanisms for abatement costs. I think it’s Coase Theorem.
Regarding socially efficient abatement costs, a method used in allocation of emission reductions among a specific industry sector was the equimarginal-principle.
The $100,000 premium for expired names could be tiered based on the constraints between variables that draw value from demand?
3-5Letter words $100,000
5-7+ Letter Words $50,000
Popular First Names $10,000
Would this achieve a socially efficient cost of mitigating name squatters?
Just this morning I have seen at least two ENS domains sell for over $20,000, so I believe this proposal was definitely necessary.
Is there a future possibility for an > (t) per voting period.
Ie for price per Eth fluctuation? For gwei?
We may reach a larger “active” voting base.
Per held balances in $ENS.
If we time a live vote for a downswing in Gwei utilization. On the network as a whole.
This is now queued for execution in 2 days.
This has now been executed.
@nick.eth Looks like there might be a display issue with the ens dapp. Says “Buy now for 0.39 ETH($984.33)” just above the descending line. Label in the chart area also says, “Starting price: $2000”
Also should have a comma, “$49145.58USD” → “$49,145.58USD” People might think it’s displaying the wrong amount.
Yup, we’re aware of it and will fix it ASAP.
this is really good solution in the short term, but to be frank, I don’t see any big problems with it even in long term
its efficient, easy to understand, does its job - why reinvent the wheel
people are already confused😐