3 ms·
> for example git could be described as a application of blockchain technology Git isn't a blockchain as it lacks a consensus mechanism. It can however be used
by p4bl0 4y ago
> for example git could be described as a application of blockchain technology
Git isn't a blockchain as it lacks a consensus mechanism. It can however be used as an append only immutable distributed ledger. If that's all you need, you don't need a blockchain.
That's precisely why consortium and private blockchains do not make any sense. Either they have a consensus mechanism that's useless and wasteful, or they're just not blockchains.
We can't call everything a blockchain otherwise it's not possible to talk about the technology in a meaningful manner. Blockchains were invented in 2008 in the Bitcoin paper. Merkle trees dates back from the 80s. It makes no sense to call a Merkle tree that happen to have the constraint of having a single child per node a "blockchain" except to ride the hype train.
Merkle trees have a lot of legitimate use cases. Blockchains only and single use case that make sense is making so-called "cryptocurrencies", which in turn has a single use case which is speculation, which is not something either useful to society or desirable for most people.
- _wolfie_ 4y ago> Git isn't a blockchain as it lacks a consensus mechanism. Isn't for example for Linux technically Linus acting as the "consensus mechanism"? Or does it have to be automated?
- p4bl0 4y agoThat's not distributed consensus it's centralized consensus, it could be automated without the need for a blockchain.