Author: StJohn Piano
Published: 2021-03-01
Datafeed Article 207
This article has been digitally signed by Edgecase Datafeed.
This article has been digitally signed by its author.
848 words - 158 lines - 4 pages

After money, a domain name, e.g. "edgecase.net", is probably the most important type of thing in the world.

Domain names are identities. Via APIs, the wealth, knowledge, and conversations of the world flow through these names. They are the busy seaports of our era.

As we migrate more and more of our lives onto digital networks, it becomes ever more crucial that identities remain fixed and solid, else long-term (i.e. meaningful) relationships become very difficult. We collapse into the endless noise of the crowd and its fleeting dreams and terrors. [0]

Every company is now a software company, or soon will be. Every organisation, every group of people, that operates in the modern world must now rely on computer networks for communication and coordination. Therefore, in order to work together effectively, every group of people will need strong forms of online identity - the stronger the better, really.

Bitcoin solves the problem of "solid money" in the computer network environment, probably as well as it can be solved. (Background: The intrinsic value of Bitcoin, Blockchain Fundamentals). There are some downsides, but everything's a trade-off. Perhaps there's a potential future version that's a bit better in some way.

So... how could we use the blockchain tech behind Bitcoin to make a new domain name system ? One that's less subject to political control ? (After the events of 2020, who among us can trust that their business won't be politically targeted, and that their domain name won't be banned / censored next year ? See The Factory Director in the U.S.S.R. by Raymond A Bauer for a worked example.)

Note: I did not say "not subject to political control". I only said "less subject".

Ideally, all transfers of domain names between parties should be blockchain-secured transactions - digitally signed and verifiable by third parties. However, we don't necessarily want to run a new blockchain that uses mining. Mining is extremely expensive, and perhaps only socially feasible for a money.

Here's my proposal, as a sequence of points:

- A domain name can be thought of as an NFT (Non-Fungible Token).

- A domain name is ideally human-meaningful.

- A BTC address can used as a base identity. It's decentralised and unique. Per Zooko's triangle, this means that it can't be human-meaningful.

- A BTC address can sign messages, e.g. transactions.

- A new blockchain could be run in parallel with Bitcoin. It would use proof-of-stake to accept new transactions. Stake would consist of BTC held in the Bitcoin address of the participant.

- Requirement to join the network: A proven amount of BTC. Perhaps this amount could adjust depending on the amount of staked capital already existing in the system.

- A transfer of a domain name from the control of participant A (with BTC address A) to participant B (with BTC address B) would be a signed transaction, published on this new blockchain.

- Every so often, e.g. once per month, the latest block could be hashed. The hash could be treated as a private key and thereby used to generate a new Bitcoin address, to which a participant would transfer the smallest possible of BTC. (The participant could use a salt, and reveal the salt only after later transferring the BTC back to another of their addresses, so as to spend only the two transaction fees.)

- The procedure in the previous point would timestamp the entire DDNS blockchain in the Bitcoin blockchain. This would take advantage of the expensive immutability of Bitcoin to make the new blockchain equally immutable, without having to run an (expensive) mining procedure in the new blockchain.

- It would probably be reasonable to run a separate token in the new blockchain. This could be inflationary - making the token into a gold (like Bitcoin) would not necessarily be helpful for the trade in domain names. Perhaps just build in a constant block reward.

- Once various exchanges add the token, it becomes relatively straightforward to turn BTC into a domain name.

- Next problem: Getting various browsers to hook into a new DNS.

- Note: Perhaps in the future there will be multiple DDNS protocols, each one roughly centered around some geographic nexus.

- Note: Could design around the possibility that Bitcoin breaks, by building the DDNS chain so that it accepts stake, and timestamps itself, in multiple cryptocurrencies.

Hm. Actually, this wouldn't have to be limited to domain names. Could be names in general. So: Not only "edgecase.net", but also "StJohn Piano" (or "stjohn_piano").

So this would really be a Decentralised Identity System (DIS). I guess a "decentralised ID" could be called a "decID" or "DID".

Conclusion: The primary incentive to participate in the network would be to "secure your own name" and protect it from political intervention (and thereby protect your business, whether that's your own job or the existence of an entire company).

[start of footnotes]


[return to main text]

[end of footnotes]