There are many, many reasons that might drive an user to abandon it's account. It can be by losing interest, losing it's keys or even by dying. Here on Hive we not only allow but even encourage people to have multiple accounts, but when dead accounts start to fill too many usernames it might become an issue.

My suggestion would be to be able to overwrite an user name from a dead account, for a greater fee that would be, lets say, 10 times the user creation fee for just to release de dead username to be overwritten by a brand new account, which would need to pay the extra account creation fee to do so.

in order for someone to claim an account to be "dead", and so, available for someone else to use, the account would have to meet some criteria:

• At least, let's say, 2 years without having anything signed under any of the account's keys or authorized proxies;
• A 28 day_{(4 weeks)} time between the request and the release of the account, so if the old user wishes to prevent it's lost, just do anything under that account. It can be a custom JSON broadcast, a 1% upvote, a profile update. Anything.

Also:

• Releasing the account's name to be overwritten won't give access to any of the previous account's assets.
  • Example: If the dead account had ~400000 HP at the time of release, the new account will start with 0.
• If by some miracle the owner of the dead account wants his old assets, he can transfer to a new account by signing the transaction with his old ACTIVE key. Any transaction done with the new ACTIVE key will only be able to manage resources from the new account.
• All mentions of the overwritten account will be considered towards the new account after it's creation. So, the old account ACTIVE key can send away stuff from the older account, by anything sent to the account will either go to the new account or give an non-existent account error if after the 28 days but before the creation of new account.
• Main Hived plugins will return new account data only, as if it was a brand new account and not an overwritten one.
  • My suggestion is for a creation of a graveyard plugin to be able to access data from the older accounts. It would be necessary to send a block number between account creation and release of the name so the plugin can track which account version is referring to, as an account would be able to be overwritten more than once. ACTIVE transactions done after the account's release(like the previous owner sending away it's resources) would still be queried referencing the old account on the graveyard plugin.
• If no new account is created under that username after being released, the old account owner would be able to resurrect it by just creating a new account with the same name BUT having the same old OWNER key.

This one might be easier to implement than the previous one:

Extended Usernames:

• The idea is to allow people to register usernames over 16 characters;
• It would cost 1 extra account creation fee for every additional character over 16.
  • I don't know how the data structure would deal with it, but it can be both stored and indexed with some database management
• Max char limit would still exist, but instead of char[16] it would be char[16][16], which is char[256].
  • That means that for every account with an extended character there would be a limit of 16 spaces with the size of a username each. Not forgetting that the additional account creation fee would be for extra character and not for extra char[16] space.
    • Example: a username using char[20] would cost 5 Account Creation Tickets to be created.
• Any transaction done by the account registered under extended char can cost more RC, in case there's additional processing needed for indexing and other stuff.
• End-user interfaces don't need to be broken in case some funny whale decides to spend over 200 account creation tickets in a huge username. Just hide the last character, replacing with a … character and only showing the full name when hovered;

That was just a couple ideas I had. I'm sorry if anything here might sound silly or impossible. I am aware about the whole costly work necessary to just plan such system for a future Hard Fork, but I just would like some feedback about it. What you liked, disliked, what is silly, what can be improved and if any of those suggestions gave you an idea for another improvement.

Thank you all for your attention.