5 ms·

A blockchain is not a particularly good solution for auditability though. A simple signed transaction log does the job just as well with much less complexity an

A blockchain is not a particularly good solution for auditability though. A simple signed transaction log does the job just as well with much less complexity and overhead. Neither approach solves the privacy issue.

lisper 8y ago
A signed transaction log doesn't solve the censorship issue either.
- silotis 8y ago
  And a private blockchain does?
  - lisper 8y ago
    No. "A currency with a ledger maintained by a TTP" != "private blockchain."
drexlspivey 8y ago
And how do you make sure that the entity in charge didn't just remove transactions from the log? Let's add a cryptographic commitment to the whole transaction set with each transaction. But then how do you order them to know which one came first. What if two transactions come at the same time? Let's split them into chunks that will be added every X minutes. And how do you make sure that the chunks aren't messed with/reordered/whatever? Let's make each chunk reference the previous one through a hash. Oh wait.
- silotis 8y ago
  Without PoW the crypto you describe does nothing to stop the TTP from re-writing history. The TTP can just re-generate the chain starting from the block they wish to re-write. With both a private blockchain and a signed transaction log you have to rely on the tx log/blockchain being published so that third parties can verify that all updates are consistent.
  - politician 8y ago
    > With both a private blockchain and a signed transaction log you have to rely on the tx log/blockchain being published so that third parties can verify that all updates are consistent. Maybe it is?
- jacques_chester 8y ago
  Hash chaining over logs predates the blockchain by a fair whack of time. If you don't mind a trusted authority cross-signing, it's a lot easier and a lot faster. You don't need the batching mechanism quite so badly and ACID databases have a pretty good track record at establishing a trustworthy ordering of events.

lisper 8y ago

A signed transaction log doesn't solve the censorship issue either.

silotis 8y ago
And a private blockchain does?
- lisper 8y ago
  No. "A currency with a ledger maintained by a TTP" != "private blockchain."

drexlspivey 8y ago

And how do you make sure that the entity in charge didn't just remove transactions from the log? Let's add a cryptographic commitment to the whole transaction set with each transaction. But then how do you order them to know which one came first. What if two transactions come at the same time? Let's split them into chunks that will be added every X minutes. And how do you make sure that the chunks aren't messed with/reordered/whatever? Let's make each chunk reference the previous one through a hash. Oh wait.

silotis 8y ago
Without PoW the crypto you describe does nothing to stop the TTP from re-writing history. The TTP can just re-generate the chain starting from the block they wish to re-write. With both a private blockchain and a signed transaction log you have to rely on the tx log/blockchain being published so that third parties can verify that all updates are consistent.
- politician 8y ago
  > With both a private blockchain and a signed transaction log you have to rely on the tx log/blockchain being published so that third parties can verify that all updates are consistent. Maybe it is?
jacques_chester 8y ago
Hash chaining over logs predates the blockchain by a fair whack of time. If you don't mind a trusted authority cross-signing, it's a lot easier and a lot faster. You don't need the batching mechanism quite so badly and ACID databases have a pretty good track record at establishing a trustworthy ordering of events.