5 ms·
Proof of work solutions to byzantine fault tolerance are vulnerable to 51% attacks almost by definition, so if there's a value to the data it needs to be backed
by 60secz 6y ago
Proof of work solutions to byzantine fault tolerance are vulnerable to 51% attacks almost by definition, so if there's a value to the data it needs to be backed by a stronger economic incentive for "virtuous behavior". Doing this outside of money is problematic.
The other option is private blockchains with limited, authenticated accounts, but then why not just use a timestamped db?
- 627467 6y agoThis. People keep wanting to strip the money from the technology in blockchain but, like many have stated: blockchain as a append only log is of little innovation or value if you have to rely on courts, lawyers and other humans outside of the technology to make it work
- quickthrower2 6y agoTrue but it does add an element of non-tamperabitiry the data. Of course if a txn is a result of some stolen key type scenario a court might be required. A correction can be added like a journal entry. But like git the history is all there and there is less to debate.
- lisnake 6y agoYes, but what if I piggyback my insignificant data on the back of some other bigger public blockchains, like bitcoin or ethereum, where cost of 51% attack is in tens of billions? That way I can get protection with their level of security.
- 60secz 6y agoBut at a significant price is complexity, latency, maintenance, and transaction fees. There are plenty of immutable append only databases, e.g. datomic. Why violate Occam's razor?
- mgraczyk 6y agoThe OP asked about hyperledger, which does not require proof of work. He/She's asking specifically about private blockchains.