Interview with the developer: What are TON DNS and Telegram Username now?

In August 2022, there was an interview with one of the developers of the TON DNS smart contract, Anatoly Makosov (tolya yanot), which is still relevant today. However, a lot has changed since then, including the launch of the Telegram username auctions — Fragment.

What are TON DNS and Telegram Username now?

TON DNS turned out great: it's well thought-through, well-designed and implemented. It's also the most multifaceted component of those launched previously, as it covers usernames for wallets, domains and subdomains for TON sites, NFT and partial burning of Toncoin emission.

In late October, the Telegram team released a username auction based on the TON blockchain. It is based on the same TON DNS contract.

Why was it decided to burn TONs paid for domains, and not to increase staking percentage?

The percentage of staking is directly related to the reward of the network validators. Validators are rewarded for keeping the network running. The more users and transactions, the more work the validators get, the greater reward they receive. This is what the tokenomics of the network is built on, and introducing new rules and relationships (such as validators receiving undeserved revenue from DNS auctions as well) is not a good idea. If there’s a need to increase validators' reward, it would be better to do so by putting forward the proposal to increase network rewards, and having the community decide by vote.

The DNS auction is made as decentralized and open as possible.

Coins from the initial sale of domains, to be precise, are not burned, but remain forever on the auction's smart contracts and are permanently removed from circulation. However, it’s easier to use the words “burn” and “burning” here.

If the funds from the initial sale of the domains had not been burned, several problems would arise:

• Who would control the proceeds? You would have to create some sort of TON DNS DAO, identify the people who would be a part of it, the format of their work, etc.
• These people would be able to take any domains, just by making huge bids on them, because at the end of the auction funds will fall into their own hands anyway.

Therefore, this option was set aside. Now the conditions for obtaining domains are equal for all: core-team participates in auctions on the same terms as other users.

Coins go out of circulation, and that always has a positive effect on tokenomics. You may argue that at the moment ~600k TON worth of domains have been sold, and that is insignificant in the context of all available coins out there, but this auction will stay forever, and a very significant amount of coins may be burned during the whole time.

Some argue that it would be better to spend the proceeds from the domain sale on registering the .ton zone with ICANN. This way, the sites would become available both at ton://dubai.ton and at http://dubai.ton. They are already working in this direction and I hope they can do it. It's worth noting that the ICANN registration fee is paid in U.S. dollars, not Toncoin.

Why did Telegram take a different approach?

Pavel Durov often says that he does not have enough money to support Telegram, and until recently, he supported everything with his own funds. Then came the paid Telegram Premium subscription, and now anyone can auction off their username.

I believe that Telegram deserves to use the Toncoins it receives for selling Telegram usernames.

There is a lot of controversy in the community about domains being too expensive, why is it so?

Domains like "life" or "casino" in the usual centralized DNS registries are not cheap either. In fact, they will cost much more than $1000 there.

If there wasn’t a minimum price for a domain, cybersquatters would quickly register all the domains with the help of automatic bots at little cost. That is why the starting minimum price depends on the number of letters — without the minimum price, you could, for example, register absolutely all 4-letter domains (26^4 = ~457 thousand) for only a few tens of thousands of TON.

The TON community is constantly growing. Setting a minimum price is fair to people who will come to TON later and learn about domains later.

It is worth noting that the minimum price decreases over time and in the end, even 4-letter domains will start at 100 TON.

The reputation of the Telegram messenger also suggests that 4-letter usernames can cost $10,000, and even 1.5M TON as time shows.

Who developed the TON DNS contracts?

Kirill Emelianenko and Anatoly Makosov are the ones who make the important Roadmap smart contracts.

They are quite paranoid about reviewing contracts and testing them. Almost all system contracts have a very high cost of failure, so it's a lot of work.

All the contracts are open source, there's a Bug Bounty program, which means that the community does the audit as well.

This year, a number of well-known companies that specialize in finding bugs and vulnerabilities in smart contracts tuned in.

How can I make sure that the DNS and Usernames smart contracts are good?

The smart contracts code is open and available on GitHub, and anyone can view it. For non-technical people, probably the main criterion is that everything works.

Telegram published a smart contract hack contest on Oct. 22 with a $100,000 prize. No one has cracked it.

How would you comment on the words of Nikolay Durov "In most cases, the persistent data of a TON DNS Resolver smart contract contains a prefix dictionary with keys equal to the defined subdomain names (with terminating zero byte) and values equal to a (HashmapE 16 ^DNSRecord)...". Why didn't you follow this scenario?

The records are stored that way, but the key size was increased from 16-bit to 256-bit to comply with the Token Data Standard and to allow different projects to simply take arbitrary strings as a key instead of coordinating the keys with each other.

For example, the Disintar Marketplace could store some of its information in user domains with the key "disintar_info" and it would definitely not overlap with other projects.

Have you thought that when there will be a full-fledged implementation of TON Sites, TON Proxy, ADNL, a lot can change? And we already have DNS contracts written. Also, in the case of any vulnerability in the child DNS contracts, the fix is likely to be impossible on the existing domains (also with NFT, tokens, etc).

Of course, everything is calculated several steps ahead.

A lot of attention is paid to revision and testing of the underlying code. The network is stable and applications don't send private keys to the backend like in Solana. To date, there have been no real hacks in TON, neither in the kernel nor in the core applications and services. The Bug Bounty program plays an important role.

For over a year and a half now, the team has been actively producing systematic releases, and has maintained a pace of several global (such as TON DNS and TON Payments) releases per quarter, on top of all other work. In fact, the pace is even faster. For example, this quarter (Q3) they also plan to launch the NFT bridge and token bridge, which were not included in the original roadmap. These are complex and non-trivial tasks, 90 days is a very short time to implement them. However, the team produces stable, well working solutions. If we consider not only the system releases, but the entire ecosystem, then news, updates or launches happen almost every day.

TON Foundation and independent development teams in TON work extremely efficiently and professionally.

The smart contract allows validators to remove or take away domains. What will happen to domains like “coinbase” or “binance”? What if the TON Foundation needs it for important tasks or purposes?

TON Foundation members buy domains on the same terms as all other users. Various exchanges, too.

In exceptional cases it’s possible to change the owner or delete a specific domain by voting. Technically, the majority of votes on the network can change not only the DNS, but any configuration of the blockchain.

DNS will run forever, so it was important to include this possibility in the smart contract, but there is no specific case where it can be applied. There must be some good reason for everyone to vote.

Thanks to the TON Foundation for open communication and answering questions.

Vudi is the author of a channel about TON @investkingyru_en and an active participant in the TON ecosystem. He is also the founder of the blockchain collection kingyTON, based on the historical events of the TON project ton.org.in. If you find errors in this post, please let Vudi know on Telegram @kingyru.