district0x - Ethereum Name Service Updates
A review of ENS project progress from Devcon3
Three of the district0x team members took a trip down to beautiful Cancún, Mexico for Devcon3. We enjoyed many presentations from the various contributors across the ecosystem, and even moreso enjoyed the networking opportunities to interface and coordinate with other teams in the space.
Some of the most exciting takeaways (not only for us, but for most of the people we talked to) came from Nick Johnson’s presentation on the Ethereum Name Service. Given our recent launch of the Name Bazaar, a secondary market for the trade of ENS domains, we’ve seen a lot of interest from our community on how these changes will effect our efforts as well as the ENS domain ecosystem as a whole.
Below we’ll summarize some of the announced changes and how they might play out in the future. You can find Nick’s original presentation on YouTube (starts at 4:05:30) and can access the slides used therein here.
Permanent Registrar Design
One of the stated goals from the ENS team since the beginning has been to iterate on the registrar design as the domain pool and usage grow in order to achieve a more equitable system for all types of participants. Nick detailed the three different user “classes” that currently exist in the ENS ecosystem.
- A regular user — someone who wants a name or two as vanity addresses for wallets. The cost to this type of user to own and retain a few names is effectively the entire cost of the deposit itself, as they will never expect to receive those deposits back if they own the names indefinitely.
- The investor — someone whose cost to hold names is weighed against the opportunity cost of reinvesting that ETH elsewhere. This opportunity cost would typically be measured in % gains, and given the low average deposit for a typical ENS domain, this results in a relatively low cost.
- The speculator — someone who was planning on holding deposited ETH anyway, and hopes to speculate on particular names only for the potential resale value (also known as “squatting”). The cost for this user is effectively zero; if their name doesn’t yield gains, they simply let it go and reclaim their entire deposit.
From the descriptions above, it becomes clear that this is sub-optimal result. The users most interested in directly utilizing ENS are disadvantaged by the current registrar design.
To counteract this, the team has decided to proceed with a rent-based model for registered ENS domains. By extracting a small, flat fee from each domain owner, the incentives for each class of user above can be more appropriately aligned. This will dissuade needless mass-squatting of names, which is ultimately a healthy change for the ENS as a whole. However, with appropriate risk assessment, a rent-based registrar still leaves opportunity for investors and speculators to participate and potentially profit from early adoption of valuable names.
Specifically for Name Bazaar and other secondary ENS marketplaces, the dynamics between buyers and sellers of domain names are given a new sense of urgency when regular fees are imposed on name ownership. Buyers on Name Bazaar can take valued names (especially newly/recently popularized ones) off the hands of the current owner for cheaper than they might otherwise be able to acquire them if de-registered and defaulted back to the public registration process.
Likewise, sellers on Name Bazaar can still balance the cost of continued rent against the potential resale value. They’re incentivized to sell names they no longer want at a discount to the perceived “true value” expected from the registration process, since it’s the only opportunity they have as owners to profit or offset the rent paid from their time owning the name.
The rate of rent will be the same across all domains, regardless of popularity (as judged by the amount of ETH deposited at registration). Rent will have a dynamic, self-regulating mechanism, whereby the more names deregister (indicating that rent is too expensive), the lower rent becomes. On the flipside, any decline in deregistration of ENS names results in a increase in the rate of rent.
All in all, these changes to the registrar process lead to a far more efficient marketplace (both primary and secondary) for domains and user across the entire spectrum, and it’s exciting to see what new challenges and alterations will be faced once this goes live
The second major announced change effects the procedure of the auction process. Currently, there’s a three step process for acquiring a name:
- The auction is started by an opening bid. At this point, a 72 hour countdown begins in which anyone can place further sealed bids.
- After 72 hours have passed, bids are no longer accepted and each bidder has 48 hours to reveal their bid. Unrevealed bids in this period could be lost forever.
- After the 48 hour period is finished, the winning revealed bid has to finalize auction in order to assume ownership over the name.
In the future, changes will be made to this process that reduce this to a two step procedure:
- Auctions for unowned names are started automatically on a rolling 48 hour basis. At this point, a 48 hour countdown begins. During this period, anyone can open a competing a bid.
- After the first 48 hours period, another 48 hour period begins in which bids must be revealed. During the reveal phase for a name, competing bids may still be placed. At the same time, any unowned names without bids have another bid period automatically started.
- The auction ends automatically at the end of the reveal phase (assuming at least one valid bid has been revealed). Any bids placed during the reveal phase are invalidated and automatically refunded. The winning bidder must finalize the auction in order to assume ownership over the name.
There are two primary advantages to this system. The first advantage comes from the removal of the “open auction” action from the procedure. This effectively lowers the required number of actions to acquire a name from 3 to 2.
The second advantage this rolling auction model provides is that because any name can be bid on at any time, no additional steps need to be taken by bidders to obfuscate their bids or hide the names they’re bidding on. From the perspective of the poacher, squatter, or griefer, all unowned names are open for bids all the time. The only reliable way to register a name is to open a bid for the maximum amount you’re willing to deposit to it, as soon as possible.
For Name Bazaar, this has little effect on our plans to push our own registration flow procedure within the dApp itself. If anything, the few simplifications make our work that much easier.
The next major announcement was (at least for our team) a massive and welcome surprise.
Through the use of DNSSEC, given a chain of trust extending from a key within the traditional DNS signed by ICANN, an owner of a domain in this system can utilize a specific ENS registrar in order to claim said name on the ENS. This, in effect, allows almost any traditional internet domain to resolve an ETH address via the Ethereum Name Service!
This system is currently functional across the .xyz top-level domain on the ropsten testnet, but in the future will be extended to operate with any RSA and SHA256 compatible domain on mainnet. With the inclusion of SHA1 support, roughly 96% of the internet’s top-level domain space will be compatible with the Ethereum Name Service.
In the long road ahead to mainstream adoption of Ethereum, integrations like this that build bridges between traditional internet infrastructure and the new ecosystems we support will provide massive benefits. By supporting most of the top-level domains on the internet, the potential user base for the ENS increases by orders and orders of magnitude.
Finally, to top things off, Nick Johnson detailed the formation of a Ethereum Name Service Foundation, a not-for-profit organization and legal entity for the development and promotion of the Ethereum Name Service.
Having a separately funded entity solely focused on the ENS will not only accelerate the rate of adoption, but will allow much clearer alignment of incentives across development efforts in the future.
We at district0x could not be more excited to see the ENS project blossom from a neat idea to a core piece of infrastructure with its own pair of legs. Each new advancement leads us one step closer to the decentralized future we all dream of.
You can find out more about the ENS project here.
For more information about the district0x Network: