AliasCoin: How To Send Money To "Satoshi" Instead of "1TE6a7tvT..."
SUMMARY
One of the chief obstacles to mainstream acceptance of cryptocurrency is the reliance on public keys to transfer funds. Such character strings are long, very difficult to memorize, non-intuitive and intimidating to new users. This proposal suggests a method by which any cryptocurrency could utilize a dedicated blockchain ("AliasCoin") to enable sending funds to a "normal" ASCII name/address such as "Satoshi".
INTRODUCTION
Imagine a bitcoin client in which you could send funds to a name like "Satoshi" or "FordMotorCompany" instead of the long, complex character strings we deal with now. This post outlines a proposal for an Alias Registry situated on a dedicated blockchain. The Alias Registry would tie a simple ASCII text string such as your name to a public key in a secure manner, enabling any other cryptocurrency to search the registry and direct transactions to the appropriate target address based on the assigned alias.
HOW IT COULD WORK
To establish an alias on the Registry users would simply submit a transaction specifying
1. The desired alias ("Satoshi")
2. One or more corresponding public keys
3. Signatures proving control of the listed key(s)
4. A registrar payment in AliasCoin.
The payment would be calculated based on how long a user wanted to register the Alias, counted in blocks. Once the Aliascoin blockchain exceeded the expiration block, the Alias would be recognized as expired, though subsequent transactions could always extend the registration duration. (It is anticipated that transactions on other blockchains such as bitcoin to an expired alias would be rejected as invalid.)
Multiple public keys on separate blockchains could be registered to the same alias. For example, "JohnDoe" could register separate keys for Bitcoin, Ethereum and Zcash in one transaction. Senders sending JohnDoe funds from each blockchain would each be able to simply direct their funds to JohnDoe, with each blockchain retrieving the appropriate public key from the Alias Registry.
The end result would be a universal Alias Registry that all cryptocurrencies could employ, once they implemented the ability to look up Aliases in the registry when they encounter a target address that is not a valid public key in their own blockchain. Users would (for a presumably trivial fee) be able to use and share easily-remembered names instead of long text strings to receive payments.
THE REGISTRAR
There had to be a catch, and here it is. I do not see an alternative to the use of a central registrar to handle Alias registrations. I am wide open to suggestions on this. The registrar would receive Alias registrations from users, verify there is no breach of a Trademark (including verifying the identity of the registering user in cases of Trademark use), and then publish the Alias registration to the blockchain. In other words, users would not directly publish their alias registrations, they would have to go thru the Registrar for verification.
Why is a central registrar body unavoidable? Because 5 minutes after launching AliasCoin without a registrar, we'd see people registering "DonaldTrump" and "IRS" and "FordMotorCompany" who have no valid connection to those names. And then we'd forever be reading about the millions of $$$ that scammers were raking in with control of those aliases.
So a global registrar body would be needed to police the system, including having the ability to cancel out invalid registrations (such as in the case of a judicial decision against a defendant in a trademark dispute) prior to their expiration. The registrar would only have the ability to delete/cancel registrations, not to alter them or unilaterally add them.
I emphasize that use of aliases would be strictly voluntary, and I'm open to ideas of how to keep the Registrar body accountable and user-friendly as possible, while protecting the public from bad actors. (SEE ADDENDUM 3) For example, the Registrar could also cancel alias registrations made by scam artists upon criminal conviction, though this needs to be explored carefully.
How would the Registrar be funded? I anticipate that the registration fees mentioned above could be directed to the Registrar to fund their operations. Such operations should be transparent and handled on a non-profit basis with full auditability via the AliasCoin blockchain, with the fee per # of blocks duration for each alias registration floating based on funding requirements. Again, I'm open to ideas how to refine/improve this. (SEE ADDENDUM 2)
A FEW OTHER COMMENTS AND OPEN ISSUES
It is anticipated that as people bought AliasCoins to purchase registrations, some would be held for future use. This would help ensure the AliasCoins would hold some value, necessary for the funding of the Registrar. I do not anticipate that Aliascoin would have significant other financial functions (and would discourage such use).
PoW, PoS? I'm open to suggestions on how mining would be conducted. If PoW, recent history with bitcoin biases me against the use of algorithms for which ASICs are available, to support decentralization.
Care should be taken to ensure that a public key used in an alias registration is solely identified with one blockchain when registered. We cannot anticipate future forks, but it is imperative that users cannot send funds to an alias and have it misinterpret the connected keys and send funds into a black hole.
For example, Mary hears John needs pizza money. She knows he uses the alias JohnDoe, so she sends him $20 worth of Ethereum using the JohnDoe alias. Only, JohnDoe has never had an Ethereum account/wallet, only one for Ethereum Classic. So he has a key in his alias registration that looks valid for Ethereum, but isn't. He doesn't even know Mary sent him funds because she forgot about it afterwards. A solution to situations like this needs to be implemented, anyone have any ideas? Is this what is meant by Replay protection? What about the thousand altcoins out there, are the keys used for each identifiably unique?
I know there are downsides to re-using public keys, and so an alias registration tied to a single public key that will be heavily used obviously conflicts with the best practice of not re-using public keys. Any suggestions to mitigate this issue would be appreciated.
Finally, each blockchain needs to know that they are reaching the correct AliasCoin blockchain, not a spoof of it, when checking alias registrations. I think this is simple enough, each transaction should include a co-signature from the Registrar proving they published it with a unique signed message, but wanted to mention it in case there are related attack vectors to defeat. (There should not be a need to verify # of confirmations in the Aliascoin blockchain for recent registrations if all transactions are being approved/published by the Registrar.)
ENDNOTE
I believe the use of Aliases could provide a big step forward in ease of use for mainstream adoption. To anyone concerned with the reliance on a registrar, I would point out that the use of aliases would be completely voluntary, and that I just don't see an alternative. Trademark infringement and scamming would explode without it.
I'm not a programmer or coin developer, so I'm just throwing this idea out for the community to consider. If it seems worthwhile and workable my hope is that a team with the right skills will step forward to start working on it. Since literally any cryptocurrency can benefit from it if developers choose to support it, I hope it proves a blessing to cryptocurrency adoption in general.
The concept of an alias registry actually dates from a project I brainstormed back in ~2014. I eventually shelved the project due to its complexity and the fact that it required ultra-high transaction capacities (far in excess of the VISA network). So I decided I'd dust it off once options like the Lightning Network became available. I only just realized this morning that the Alias concept could be applied completely independently and the benefits could be shared with everyone right now.
If this idea bears any relationship to other proposals out there that have already been explored or even implemented, I apologize for being unaware of them. It's not my intent to waste anyone's time.
Throwing it open for discussion...
ADDENDUM 1: WHY A BLOCKCHAIN?
It occurs to me that people may ask, if you are going to need a centralized registrar, why even bother with a blockchain? Why not just trust them to keep the alias-key pairs on a private database?
The answer is that such databases can be hacked, and there would be tremendous incentive to do so if the use of aliases become widespead in cryptocurrency. The most secure method of storing information is on a blockchain. It is hackproof and unalterable apart from defined transaction types. The registrar would need to exercise due care to keep its own keys safe (presumably utilizing some form of multi-sig).
Perhaps it might be necessary to implement a feature by which the public keys used by the Registrar could be "invalidated" on the blockchain - that is, if any given key used by the Registrar was lost/stolen/compromised, the Registrar should be able to send a multi-sig transaction to the blockchain that invalidates the use of the compromised key in any future transaction on the blockchain. Thus even if an attacker obtained several keys over time from the Registrar, it would be unable to use them if the hacks were detected. (Recommendation: The Registrar could make it a practice to invalidate keys it uses at a regular frequency, introducing new ones, to mitigate the risk of undetected thefts.)
Also note that the Registrar should not be able to spoof Alias registrations. Alias registration submissions from users should include a signature (or signatures) proving that they control each of the keys listed in their alias registration. This would improve trust in the use of a registrar, since they can only delete alias registrations (such as in a case of trademark infringement), never alter or add them unilaterally.
ADDENDUM 2 - MINER-FUNDED REGISTRAR?
This is another rough idea I worked out in my original project. The idea here is that registration fees would be paid to the miners (like in bitcoin) rather than the Registrar. To fund the Registrar, miners would have the ability to generate Aliascoins with each block that would be paid to a public key issued by the Registrar. The funds going to the Registrar would be in ADDITION to the regular miner payout per block, and could be specified within some range by the miner.
Thus miners believing the Registrar is overfunded could cut back funding, and those believing it underfunded could increase funding, both with no direct impact to the miner. Miners would be incentivized to not overfund the registrar because it would dilute the value of their own mined coins, and would be incentivized not to underfund it because it would risk the smooth functioning of the Registrar that upholds the value of the Aliascoin project and the value of their own coins.
Whether this is workable and would achieve the right balance needs to be carefully explored. But I thought it an idea worth mentioning, because...
ADDENDUM 3 - A REPLACEABLE REGISTRAR?
This is more speculative, but bear with me. What if it was possible in principle to replace the Registration body with a competing organization in one fell swoop through a democratic process? As an anology, what if the economic majority could vote on whether they preferred Core or Bitcoin Unlimited as being in control of the Bitcoin protocol?
Here is the proposed method: The Aliascoin protocol would include a voting mechanism (such as Dash, etc. employ), in which each coin counts as one vote. The more coins you have, the more voting power. Votes would be tallied based which public key you support - the one of the current Registrar, or one published by a competitor.
Depending on the exact rules for determining a victory (plurality, majority, majority after run-off, etc.), the funds allocated by miners would automatically go to the public key of the winning Registrar body. Moreover, only the winning body holding that public key would be able to submit registrations to the blockchain, effectively acting as the official Registrar. The keys employed by a Registrar driven from power would no longer be accepted for such duty. (I leave it for developers to determine the exact method to do this and how it could be made secure.)
Why on earth might we want the ability to change the Registrar?
1. It is anticipated that the Registrar would also take a lead role as developer for the blockchain. In case of serious conflicts such as the Core/BU debate this would provide a mechanism for resolution of the debate.
2. Insofar as the Registrar is made up of real people in real locations, the Registrar body could be subjected to intimidation or other coercive acts, or even physically replaced, by local/national authorities or organized crime. The voting mechanism would enable the community to disenfranchise such a compromised or corrupted Registrar, replacing it with a fresh body that obtains the most support.
In any event, the Registrar body would need to uphold two linked ideals: The protection of properly trademarked material to the best of their ability, and the prevention of criminal exploitation of aliases.
