3 ms·

For many of them you only need an append only ledger not a full blockchain. And you still have to deal with the oracle problem. Is my Picasso real? - If I trus

For many of them you only need an append only ledger not a full blockchain. And you still have to deal with the oracle problem. Is my Picasso real? - If I trust the entity that put it into the blockchain, then I believe so. Wait, why did I need a blockchain for that?

shawabawa3 9y ago
> If I trust the entity that put it into the blockchain, then I believe so. Wait, why did I need a blockchain for that? 99.9% of these "trust ledger" ideas have this problem.
- tonyjstark 9y ago
  That's why I wonder why it gets so hyped. Many of those ideas get sold as if you can guarantee correctness of everything, solve every problem once and for all. But nobody thinks about the problem that you can still manipulate the data before it is in the blockchain.
dreit1 9y ago
Trust isn't neccesarily the only solution that blockchains solve. Lets consider a simple example, login credentials. Your login credentials exist on many different centralized databases, which a service provider is responsible for managing. This leads to increased possibility that one of the providers messes up (identity breach) and leads to an unnecessary amount of data replication across databases. Why do I need credentials for every single different service? Instead why don't we store these login credentials in a blockchain so that many providers can access these IF you let them, in a cryptographically secure way. You PROVE your identity on the blockchain without actually letting them store identity information on their database. A blockchain is basically a shared resource pool where a user chooses who can access the data that belongs to them whether it be assets, identity, or other important data. Service providers no longer have to be responsible for it and the user is put back in control. This is a much saner way of storing important things on the web
- madeofpalk 9y ago
  Couldn't then a bad actor from obtaining >50% of computing power behind it and manipulate it? It always seems like you have to trust someone which still begs the question - what's the point?
  - dreit1 9y ago
    Im all in on dBFT, so I believe in the elected delegates maintaining said distributed database. Similarly to how governments maintain roads, I trust the elected parties to not corrupt the database. I think complete trustlessness is a false dichotomy. Im ok with trusting certain parties in maintaining distributed "value databases" as one might call them} Edit: false dichotomy is bad phrase, idk "chasing the wrong goal?"
    - annabellish 9y ago
      The problem is that a blockchain which does not allow third party miners gains nothing from being a blockchain, because it can be a traditional database without losing anything at all. A blockchain which does allow third party miners is open to attack without obscene resource investment.
      - sharemywin 9y ago
        as far as I understand it miners only bootstrap the system eventually transaction fees run it.
      - dreit1 9y ago
        In dBFT the attack vector is a collusion attack between elected delegates. In practice I do not see this being any more likely than a 51% hashpower attack. However most delegated systems are still young. The only reason people believe that 51% attacks are impossible is because it has worked for 10 years. If delegated systems work for the same amount of time, I don't see why they wont get the same reputation. I think Dan Larimer gives a good argument on the situation. https://steemit.com/eos/@dan/response-to-vitalik-buterin-on-eos https://steemit.com/eos/@dan/response-to-vitalik-buterin-on-... Edit: His Conclusion is something many people should consider, when outright rejecting delegated consensus. "EOS is designed around far more realistic assumptions and logic and achieves scalability by avoiding the dogmatic fallacies promoted by Bitcoin and Ethereum maximalists."

shawabawa3 9y ago

> If I trust the entity that put it into the blockchain, then I believe so. Wait, why did I need a blockchain for that? 99.9% of these "trust ledger" ideas have this problem.

tonyjstark 9y ago
That's why I wonder why it gets so hyped. Many of those ideas get sold as if you can guarantee correctness of everything, solve every problem once and for all. But nobody thinks about the problem that you can still manipulate the data before it is in the blockchain.

dreit1 9y ago

Trust isn't neccesarily the only solution that blockchains solve. Lets consider a simple example, login credentials. Your login credentials exist on many different centralized databases, which a service provider is responsible for managing. This leads to increased possibility that one of the providers messes up (identity breach) and leads to an unnecessary amount of data replication across databases. Why do I need credentials for every single different service? Instead why don't we store these login credentials in a blockchain so that many providers can access these IF you let them, in a cryptographically secure way. You PROVE your identity on the blockchain without actually letting them store identity information on their database. A blockchain is basically a shared resource pool where a user chooses who can access the data that belongs to them whether it be assets, identity, or other important data. Service providers no longer have to be responsible for it and the user is put back in control. This is a much saner way of storing important things on the web

madeofpalk 9y ago
Couldn't then a bad actor from obtaining >50% of computing power behind it and manipulate it? It always seems like you have to trust someone which still begs the question - what's the point?
- dreit1 9y ago
  Im all in on dBFT, so I believe in the elected delegates maintaining said distributed database. Similarly to how governments maintain roads, I trust the elected parties to not corrupt the database. I think complete trustlessness is a false dichotomy. Im ok with trusting certain parties in maintaining distributed "value databases" as one might call them} Edit: false dichotomy is bad phrase, idk "chasing the wrong goal?"
  - annabellish 9y ago
    The problem is that a blockchain which does not allow third party miners gains nothing from being a blockchain, because it can be a traditional database without losing anything at all. A blockchain which does allow third party miners is open to attack without obscene resource investment.
    - sharemywin 9y ago
      as far as I understand it miners only bootstrap the system eventually transaction fees run it.
    - dreit1 9y ago
      In dBFT the attack vector is a collusion attack between elected delegates. In practice I do not see this being any more likely than a 51% hashpower attack. However most delegated systems are still young. The only reason people believe that 51% attacks are impossible is because it has worked for 10 years. If delegated systems work for the same amount of time, I don't see why they wont get the same reputation. I think Dan Larimer gives a good argument on the situation. https://steemit.com/eos/@dan/response-to-vitalik-buterin-on-eos https://steemit.com/eos/@dan/response-to-vitalik-buterin-on-... Edit: His Conclusion is something many people should consider, when outright rejecting delegated consensus. "EOS is designed around far more realistic assumptions and logic and achieves scalability by avoiding the dogmatic fallacies promoted by Bitcoin and Ethereum maximalists."