0

I'm not sure if this is the right place, but I have an idea for a database kind of like a wiki, that would NOT have editors, and would have the following requirements:

  • Any user can start with using their real name or a pseudonym
  • Anyone can later switch to using their real name (no reversals to a pseudonym or changing pseudonyms)
  • Identity is assured using some combination of a private/public key system and hashing (data is hashed with the content and the key of the user that created it).
  • Any user can add data and/or comments to anything (but cannot delete another users data).
  • A user can "delete" their own data, which just creates a new version of the same "chunk". This is visible and browsable by anyone.
  • Any user browsing/reading the database can filter the raw data itself by any desired means. For instance: only show data from real-named users, only show data that hasn't been changed, only show data that has been verified or confirmed by X number of other real-named users, etc.

I see some issues which might be fixable or unsurmountable:

  • Verification of real name. This could involve some trust system, where multiple (unrelated) parties have to sign off on who you are (not on the content of your data, but just your identity).
  • For two people that share the same name, there will need to be some additional name field, maybe an incrementing number.
  • If you lose your private key, how do you post again? Could there again be a way where a quorum of the verification parties ID the real you, as you, and your private key can be returned? (this sets up issues with storing the key)
  • Preventing masquerading for the pseudonym accounts (maybe data can be filtered by only those who will actually sign their name), so that the hackers/spammers don't overwhelm the system with bad or fake data, that is seemly supported by a huge body (of fake people).

My question is, where do I go to learn about how this might be implemented, or contacts I can ask about the details that need to be fleshed out or refined?

Adam Jones
  • 101
  • 1
  • 1
    This might be more appropriate here: http://security.stackexchange.com/. – Yuval Filmus Feb 23 '17 at 21:21
  • This question might be overly broad. Designing an entire system might be beyond the scope of a single Stack Exchange question. Can you break this down into multiple smaller standalone questions? You might want to learn about authentication and public-key cryptography. You should think about what the threat model and trust model is (is the server trusted to preserve people's privacy or not?) and what the security requirements are. Asking for contacts or recommendations of people you can talk to is beyond the scope of Stack Exchange. Implementation concerns are beyond the scope of CS.SE. – D.W. Feb 23 '17 at 21:27
  • 1
    Usually I don't post such broad questions. However, one doesn't know what one doesn't know. And this was more of a reaching out for ideas as to where I might look to find out more, and communities which might be open to commenting on it. – Adam Jones Feb 23 '17 at 21:29

0 Answers0