Permanent Registar Proposals - Overflow Funds, Fees & Rent, 3-6 Character Auction Start



That sounds like a nightmare to me. Adminstratively and legally.


That’s a fair point, but I think it assumes more ‘flatness’ in the namespace than is really the case; the price we set needs to be above the cost someone would willingly pay for the ‘worst’ name of that category. If we set the price at $10/year for 3 letter domains, I’m concerned we’d just see every 3-letter domain snapped up and resold; for token names alone there’s a huge market.

I really want to step aside from this as an operating concern. I’m fairly confident ENS can find grant funding if needed; while it’d be nice to not have to depend on that, I don’t want that to influence pricing policy.

Can you elaborate?


I dunno, would you pay $10/year for jkbx.eth? :grin:

Looking at coimarketcap’s top 100 tokens 27% of them use tickers of four or more characters, so attempting to lock down every <ticker>.eth by squatting every 3-letter name will just cause people to move to slightly longer tickers (squatting all 4-letter domains at $10/year would cost $4.5MM so seems unrealistic to me).

On a slightly broader point, we always discussed the idea of ENS providing most utility with subdomains. Sticking with the example of tokens and coimarketcap as a recognised and trusted provider of information, it’s easier for the average user to trust <ticker>.coinmarketcap.eth than just <ticker>.eth, and for the untrusting user to verify ownership of coinmarketcap.eth and then access multiple subdomains rather than verifying ownership of each individual domain. If this model holds then there really isn’t much to value one name over another as it’s whatever the service provider picks and publicises (coinmarketcap.eth, token.eth, cmc.eth, tokeninfo.eth, etc.)

I know that there is always going to be a tension between legitimate use and squatting, but I do worry that high rental will both discourage ENS adoption and open up the ENS foundation to attacks for price gouging (in addition to the concerns I already raised about logographic and similarly compact languages.


Here is a very good article you can review. Especially if you are interested in adjudicating marks yourselves.

I am only going to put this little bit here below . Ask if you want more:
Rule 4 (a) mentions applicable disputes in the events wherein:

The domain name is identical or confusingly similar to a trademark or service mark in which the complainant has rights; and

There are no rights or legitimate interests in respect of the domain name; and
The domain name has been registered and is being used in bad faith.
Rule 4 (b) explains the evidence of registration and Bad Faith use.

Direct or indirect laws may be invoked to curb the menace of cyber squatting and domain name disputes through adjudication mechanisms in place but is our domestic and international legal system proficient enough to actually enforce the same efficiently?

© Uniform Domain Name Dispute Policy (UDRP)

In cases of cyber squatting, cyber piracy or bad faith registration, the complaining party can invoke a special administrative procedure before the WIPO Administrative and Mediation Centre to resolve the dispute. Under this procedure, neutral persons selected from panels established for that purpose would decide the dispute. The procedure takes approximately 45 days, costing about $1000 in fees to be paid to the entities providing the neutral persons and is handled mostly online.

The Rules clearly lay down:

Under Rule 4(k) the parties to the dispute can also alternatively go before their domestic judicial courts to resolve the dispute or to contest the outcome of the WIPO Administrative Panel proceeding within 30 days of such impugned decision. Every ICANN accredited registrar has agreed to adhere to the dispute resolution policies that ICANN adopts under its established consensus procedures.

Administrative proceedings, fees, disputes/ litigation, remedies, availability of court proceedings, cancellations, transfers, changes, transfers during disputes and policy modifications are all clearly laid out.

They define various terms like complainant, ICANN, mutual jurisdiction, panel, panelist, party, policy, provider, registrar, registration agreement, respondent, reverse domain name hijacking and supplemental rules.

They lay down the method and the manner of communication as well as the form and format of a complaint. They specify when and how a complaint is to be notified and to inform the complainant and the respondent if the complaint is administratively deficient for the correction of which 5 days are given.

The time period to file a response to the complaint is 20 days of the date of commencement of the administrative proceeding.

Manner of appointing panelists for a dispute and in what manner the fees are to be paid by the parties are clearly explained. In all cases the fees are to be borne by the complainant except where the respondent has made a request for a three member panel in which case the fees shall be shared equally by both parties.
The parties are banned from communicating unilaterally with any panelist. The Rules are not to be amended without consent or prior notification of ICANN.

Under Rule 4 the disputes are specified for which mandatory administrative proceedings are held. The proceedings are conducted before one of the administrative-dispute-resolution service providers listed at

Rule 4 (a) mentions applicable disputes in the events wherein:

The domain name is identical or confusingly similar to a trademark or service mark in which the complainant has rights; and

There are no rights or legitimate interests in respect of the domain name; and
The domain name has been registered and is being used in bad faith.
Rule 4 (b) explains the evidence of registration and Bad Faith use.

I wont bother getting into what Rule(b) entails because those are questions of fact that only juries and judges should make.

The WIPO Arbitration Rules are applicable to the arbitration proceedings wherein the powers of the tribunal, schedule of fees, conduct of proceedings, awards, costs etc. are expressly provided for. According to Article 46 of the Arbitration Rules interim measures of protection and security for claims and costs can also be granted by the arbitration panel for effective remedy.

Is this really what the ENS group wants to spend their time doing?


Or xkcd.eth? :wink:

The point isn’t to have the entire namespace registered - quite the reverse!

I think that’s unlikely - I don’t think people will select their ticker based on ENS .eth availability. But if it’s affordable, I think squatters will grab all 3-letter domains just in case. I think that we should increase their cost because that discourages speculating on a very limited resource.

I agree - that’s why I set up thetoken.eth, for instance. But short names are always going to be attractive.

I’m generally persuaded by your arguments that the pricing should be ‘flatter’ than what was initially proposed - but I still think making it entirely flat would be a mistake.


The above information came from here:


Definitely not, but there are simpler approaches. Let’s spin this out into a separate thread.


I tried to post this directly to moderators but I am limited to posting it here. However; may be useful to everyone who is interested in system development.

I suggest a working Group be set-up under True Name or ENS for the purpose of researching and reserving specific names. These names will be rented just the same way everyone else will rent names in the system; no early reservations and the Group will be subject to the same market rate pricing set up by ENS.

The reserved names will be used for the sole purpose of providing free sub-domain to users to help gain awareness to the ENS and wider Ethereum ecosystem.

While this functionality is available today, there has not been a concentrated effort to isolate any specific name and promote it in any significant way. For example, something like Friends. eth or MyCity. eth or ItsCool. eth may lend themselves to “in-house” development.

The examples above are just off the top of my head so apart from the obvious, there are dozens of examples of what would be worthy for “in-house” use and promotional purposes.

One great way to build Project awareness is to give something away for free that will lead it’s users to be just a few clicks away to a Web3 interface.

The best part it is there is nearly zero capital overhead required and the end-user may actually end up using Johnny.IsCool.eth as his root for every on-line transaction because it’s connected to his identity and SHA3 account.


Hello. Makoto from ENS team here.

We are planning to port to ens manager so that people can buy/sell subdomains of their own from our manager. The contract is here. Even now, you can search what subdomains are available here . Some of subdomains are even offered for free.

Would picking selected reserved name add significantly more value as compared to this open approach?

If we officially pick one or two selected domains for free and give special recognition, we are vulnerable to the same squatting attack which we are introducing the rent to begin with while damaging this open market approach I think.


Yes, I think it would add significant value. The list you mentioned has marginal names at best. There should be a genuine marketing plan around the selected name(s).

Have the working group keep it secret. It would only be a few target names. Less than ten would be selected to go through the bid/buy/rent process.

It would also add to user engagement and hopefully adoption.


What I meant is that if we support only a handful of “special” domain (eg: e.eth) and give a genuine marketing effort, squatters can squat on each subdomain (eg: facebook.e.eth) just like they try to squat on the second level domain because we are already marketing second level domain are special. If we don’t give any special treatment to any second level domains, then squatters cannot squat because the number of free subdomains are infinite.


Would there be some reason that someone would do that? Would it be just to clog the system and make it inoperable?

The sub-domains can’t be re-marketed, correct? Isn’t the root domain owner set up to be “overseer” and all “approvals” are required for sub-domain ownership changes?

Maybe a multi-sig requiring at least one registered registrar to OK the ownership transfer?

Someone may try to do private sales/resales exchange in Ether for ownership transfer outside of the naming system, i think sub-domains can still be subject to programmable restrictions?

At least I hope they can be.


Currently no but if ENS officially appoints the special domain and tries to make it more valuable than others, then it maybe. In reality I do agree that no one would do that apart from annoying ENS team.

In fact we actually got ensdomains.eth and may offer subdomain under the domain for free (I think @lyricalpolymath is advocating for it ), but not sure if we want to create a working group to pursue what names to be reserved for the purpose.


No, the owner of a name can transfer it as they wish. The owner of the parent name can also reassign it. I wouldn’t really want to try and enact restrictions on transfer, because they’re a hassle and also fairly pointless - someone can just register each name with a new holding contract, and transfer ownership of that to the purchaser.


Dear all,

I’ve been following ENS since its very beginning as one of the silent observers. Now that the permanent registrar is arriving, it feels like any additional input from the community could be helpful and I would thus like to share my personal point of view, although some aspects might not be new.
Before I do so, I would like to pay my greatest respects to Nick and the entire ENS team. You do an amazing job despite the fact that there are still a lot of tough questions to be answered.

Overflow Funds:
The use of overflow funds by the ENS team/True Names LTD is not only justified, it is crucial for further development of ENS as a core ethereum functionality that cannot depend on donations or a rising price of ether. The main question here is how this is communicated: ENS should make it very transparent how overflow funds are used, that salaries are not excessive and remaining funds are used to foster this important part of our ecosystem. If the funds help ENS to succeed, everyone wins: Users (clearly the most important group) can only make use of ENS if it succeeds and even squatters should not bother to pay their contribution as otherwise their names are worth zero.

Fees and Rent:
It is easy to understand why you labeled shorter names with higher prices, although the factor of 4 is of course arbitrary. (The factor should rather be 26 if you consistently wanted to account for the scarcity in case you had e.g. 26 different characters for an ENS name - just a thought, I don’t want to propose this.)

In fact, I am not so concerned about the yearly prices for the 3 and 4 character names. I am rather concerned about the longer names of the top owners (

Why? Scripted mass-registering frustrates the everyday user when any possible name is taken - I am not sure whether subdomains solve this issue as they are usually long and quite unattractive. The fee for 6+ characters (no matter if 10$ or 5$ per year) is thus absolutely needed.

In contrast to the longer names, I, as a user, can understand that short generic names (bank.eth, money.eth, …) are interesting names for a lot of people and it would not stop me from using ENS if I was not able to register and use one of these premium names. I also found the idea of using DNS as a reference for brand names interesting but see too many open organisational/governance/legal questions.

Most of the proposals I have read here aim to come up with a fair model, which is a remarkable and noble approach. However - and I don’t want to sound too pessimistic here - the word “fair” is extremely subjective. What is “fair” to one person may mean the exact opposite to another person. You could make short names extremely expensive which may then be fair for whales, the standard capitalistic approach. I personally would favor a first-come-first-serve system if we found a registration mechanism that cannot be automated/scripted and does not complicate the user experience. Is it not possible to have a straight forward registration process via the ens manager with some sort of (very) extended captcha and an explanatory text why this is necessary?

Additional thoughts:
Window between permanent registrar deployment and 3-6 character auction start: Here I don’t have a preference.

Harberger tax: Just no. Highly interesting theory and proposals but no company, no institution will ever use ENS if they cannot plan and make sure they keep their name.

Adoption: As in many other blockchain applications, I see the lack of privacy as a hindrance for adoption. If I have a wallet and an associated ENS name and want to accept a payment from a another person. He/she will forever know my address and my account balance. We can find arguments against that, but let’s be honest - no one wants that.

Besides these issues, I am optimistic and looking forward to see ENS succeed.


Thanks for your thoughts, @interested-in-ens.

There’s not really any way to enforce this - and even if there were, we’d surely see a rush of people “playing the lottery” for short expensive domains just so they can resell them to others at higher prices anyway.

There’s also two issues here: initial allocation of short names, and ongoing cost. The current goal is to allocate the former using an auction, with a fixed rent cost thereafter. The rent cost is intended to incentivise giving up a valuable name if it’s not being used.

Bear in mind that this is an issue with Ethereum itself - you’d have the same issue without ENS, giving someone your account address.


Several good points have been made regarding pricing, especially around the relative scarcity of 5+ character names, and the current cost of registration. In response to this, I’ve adjusted the suggested pricing:

  • The base price is $5/year instead of $10/year
  • 5 character names are now the base cost of $5/year

Another issue that hasn’t been given much thought is the potential for a pre-transition land rush to get cheap names: If you can register a name immediately before the new registrar is deployed, and migrate it immediately after, you get a year’s free registration, but also don’t have to lock your funds up. To combat this, I’m suggesting enforcing a minimum of 6 months between registering a name in the current registrar and migrating it to the new one.


couldn’t we imagine that you simply pay the registration fee when you claim it in the registrar in the first place? and contextually you get your locked ETH back.
and actually put a time limit of 6 months for people to migrate their domains to the new registrar (or releasing them)?


Doing this is awkward, because it introduces a dependency in the new ‘base’ registrar code on pricing information.


Are the current name registration clients going to make people who are registering names during this “6 month” period aware that these names may not transfer to the new registrar?