Preserve valuable DTag from being registered

[RiccardoM]

Since Desmos will allow any user to register their DTag from day one, this might cause some users to abuse this freedom and register multiple valuable DTags (eg. short ones or the ones associated with real VIPs). This has already happened in centralized social networks were very short usernames are now worth a ton of money.

I think we should define some ways to prevent this, to make sure that users cannot register those DTags only to re-sell them in the future. Me and @g-luca have thought that in order to solve we could:

1. Force the length of the registered DTag to be at least 6 characters long.
   This way we can preserve short ones and maybe sell them later with some kind of auction.
2. Set a list of DTags that cannot be used.
   This should help us making sure that none uses tags like elonmusk, rogerfederer etc. This list can be present inside the x/profiles parameters and updated through governance proposals.

To have the best prevention against this kind of economic attack, we should implement those methods together so that all tags with less than 6 characters + all VIP tags can be preserved and sold later.

Note
In the future, the DTag auctions can be used to gather money for different social events (eg. like what Starname did selling short names to raise funds against the COVID pandemic).

Please let me know what you think @kwunyeung @bragaz @dadamu

[leobragaz]

This make a lot of sense, considering that a person can have multiple desmos addresses and profiles possibly collecting a lot of valuable DTags. It's worth to explore DTag auctions, maybe managed by a smart contract?
Btw, if we force DTags to be at least 6 chars long to preserve those short ones, we need to rethink a little the profiles params, cause there could be some problems if for example we have the length param set to 6 and then we got some DTags with 3 chars. This can easily break the invariants if we don't pay attention.

[kwunyeung]

Oh haha! I suggested this when I first mentioned the idea of DTag but you guys thought it would affect the free market and convinced me not to have this feature. The example I brought out was Handshake.org. They reserved 100,000 domain names form the top 1m visited websites on Alexa. They will wait until the owners of the original domain names claim those names. If you search those domain names, e.g. apple, on Handshake, you will see this.

They even don't have an estimated release date of the domain name. It won't be available for bidding. And this apply to desmos/ on Handshake as well.

I do agree to have such arrangement. The problem is how we generate the list of preserved DTag. How to get popular usernames from Twitter, Facebook, Instagram, Youtube, Reddit?

For DTag auction, I do would like to see if this can be done with smart contract. If this will be managed by SC, then the transfer can only be available via SC but not original messages like now. Also, if we make it like auctions, we can add NFT elements in it. For every exchange, the original minter of the DTag could have a share of that exchange? We also mentioned about taking a portion of the exchange amount as community tax.

[RiccardoM]

I would like to get @bragaz and @dadamu to work on this question specifically:

how we generate the list of preserved DTag. How to get popular usernames from Twitter, Facebook, Instagram, Youtube, Reddit?

Could you please try and see if there is any existing list of most popular names used? What I would like to have is a list of usernames used by VIPs

[leobragaz]

looking online I've found out that twitter has its own lists of verified accounts divided in categories:
https://twitter.com/verified/lists.
Unfortunately I haven't got the same luck with others SN. There's seems to be no other lists of verified accounts.
Talking about companies we can possibly get those from the S&P 500 and any other known indexes.
I will research more to find out if there's a feasible way to collect these data

[kwunyeung]

Hm… that list doesn’t make any sense to me. Seems it’s only a list that maintained by @verified which are not verified Twitter accounts.

[dadamu]

I have found the site gathering the most followed profiles rank here. I suppose we can refer it to build the list.
https://www.trackalytics.com/

[dadamu]

From my side, Avoiding users registering famous or popular names in the anonymous network doesn't make sense to me. We might be able to set them to the list but it would include only a little part of them, and the users having the same name in the list must change another name since the list will reject them. The idea sounds that the list protects the famous people and the DTag economy system. However, the DTag troll still can register other popular names. Say, Chen is the popular last name in my country, the names in English are duplicated easily so the troll can register a set of profiles having DTags like paulchen, davidchen, robertchen and etc.

Honestly, human-readable name should not be unique in my opinion. In my experience, allowing duplicated name in the system is the trend. Many of games applies the way to assign the unique id and let users set their name. In this model, wallet address is the unique id already. Why don't we simply let DTag duplicated or set DTag has random number suffix like paulchen#7777 as Discord does?

[dadamu]

I agree relying on third party platform is the better way to the issue. The trust should be built by the guarantor who stakes a lot of resource. Initially, relying third party platform is the best selection for the case since they built for it with a lot of resource and the proof is already there.

For the Discord usernames sample, I consider that the troll would not appear if occupying a name is too costly. However, as you said, the inner process to deal with it is complicated. And true, It might not be so much value if the profile creation/saving is not costly.

[leobragaz]

I agree with @kwunyeung when he said that the verified lists on twitter are useless for our purposes, plus there are no feasible ways to build a good list like we intended to. I think the best way to handle this would be by using the third-party authentication as a proof as you guys stated, since there's no way to trick that system. I would also avoid to add further complexity to the DTag system.

[kwunyeung]

I prefer we skip this limitation if we cannot find a good solution to generate the list of existing accounts. And I agree with @dadamu that this feature will become a protection to the existing famous people. I have rethink the same situation on Handshake.org. If they didn’t reserve desmos/, I have already registered it. Why they reserved it for the owner of desmos.com but not us who would sincerely appreciate the decentralized domain name?

I registered elonmusk to demonstrate we don’t trust by only looking at a username. Also if there is a trend of registering popular DTag may become a good marketing tool as we will see a lot of DTag registrations in short time. It may help the future DTag exchange marketplace and benefits to the community as all trades should be taxed and contribute to the community pool. Active trading of DTag will be something we would be happy to see.

To protect some of our early supporters, how about we let teammates, validators and contributors to register their DTag on testnet and we migrate them on mainnet genesis?

[RiccardoM]

To protect some of our early supporters, how about we let teammates, validators and contributors to register their DTag on testnet and we migrate them on mainnet genesis?

I don't think this would be fair either. Also, inside the testnet the DTag params are different from what will be in the mainnet. Inside the testnet you are allowed to register a DTag if that is long at least 3 characters, while in the mainnet we will be setting it to be at least 6. I think if everyone starts from the same base it would be more fair

[kwunyeung]

Okay. Let’s start mainnet with everyone the same.