4 ms·

Any blockchain which universally stores media forever puts itself in the category “economically unfeasible/unsustainable.” People pay for that storage and the l

by mw888 5y ago

Any blockchain which universally stores media forever puts itself in the category “economically unfeasible/unsustainable.” People pay for that storage and the large and distinct nature of media makes it easy to single out prune. Worst case scenario I replace that block with illegal media with its original block header and no data. The cost to sustain an illegal media attack such that consecutive blocks contain such data is unsustainable and far easier to ameliorate than a 51% attack. According to this $76k per gigabyte on Eth. https://ethereum.stackexchange.com/questions/872/what-is-the-cost-to-store-1kb-10kb-100kb-worth-of-data-into-the-ethereum-block https://ethereum.stackexchange.com/questions/872/what-is-the...

Grimburger 5y ago
> Worst case scenario I replace that block with illegal media with its original block header and no data How do you alter a block without altering the headers of every other block afterwards? Unless you are capable of second pre-image attacks on modern cryptographic hashes that's simply not possible. What you have is a database, not a blockchain.
- nemo1618 5y ago
  You can release a hardfork, such that all participants agree to treat the offending block as a special case. No different from the Ethereum DAO fiasco.
  - capitalsigma 5y ago
    Saying "just hard fork it" is kind of admitting defeat, no?
    - nemo1618 5y ago
      Yes, personally I would not recommend this approach, just wanted to clarify how such a thing could be achieved.
  - webmaven 5y ago
    > You can release a hardfork, such that all participants agree to treat the offending block as a special case. No different from the Ethereum DAO fiasco. How do you get everyone to agree to fork each time someone stores or identifies offending material (or a pointer to such) on chain? At least in the Ethereum DAO case, everyone had a vested interest in switching to the fork. You would need a regularly scheduled "Hard Fork Thursday" or similar to effectively address the issue, and that's quite unlikely as this introduces new vulnerabilities.
breadbreadbread 5y ago
What I get from this is that all web3 promises are smoke and mirrors. ok your photos arent stored on the blockchain... the blockchain just stores your image URLS.... so then if there is still a traditional repository for content off the blockchain, you now have two points of failure instead of one.

by mw888 5y ago

Grimburger 5y ago

> Worst case scenario I replace that block with illegal media with its original block header and no data How do you alter a block without altering the headers of every other block afterwards? Unless you are capable of second pre-image attacks on modern cryptographic hashes that's simply not possible. What you have is a database, not a blockchain.

nemo1618 5y ago
You can release a hardfork, such that all participants agree to treat the offending block as a special case. No different from the Ethereum DAO fiasco.
- capitalsigma 5y ago
  Saying "just hard fork it" is kind of admitting defeat, no?
  - nemo1618 5y ago
    Yes, personally I would not recommend this approach, just wanted to clarify how such a thing could be achieved.
- webmaven 5y ago
  > You can release a hardfork, such that all participants agree to treat the offending block as a special case. No different from the Ethereum DAO fiasco. How do you get everyone to agree to fork each time someone stores or identifies offending material (or a pointer to such) on chain? At least in the Ethereum DAO case, everyone had a vested interest in switching to the fork. You would need a regularly scheduled "Hard Fork Thursday" or similar to effectively address the issue, and that's quite unlikely as this introduces new vulnerabilities.

breadbreadbread 5y ago

What I get from this is that all web3 promises are smoke and mirrors. ok your photos arent stored on the blockchain... the blockchain just stores your image URLS.... so then if there is still a traditional repository for content off the blockchain, you now have two points of failure instead of one.