3 ms·
In a sense, git is “blockchain technology” that predates the buzzword: every commit is cryptographically hashed, including the hash(es) of its parent(s). This
by kd5bjo 7y ago
In a sense, git is “blockchain technology” that predates the buzzword: every commit is cryptographically hashed, including the hash(es) of its parent(s). This lets anyone verify all the involved commits back to the beginning of the chain, and is what enables the 3-way merge algorithm without a centralized server.
What it doesn’t include is any kind of proof-of-work or proof-of-identity, which some people might consider necessary for something to be “blockchain technology”.
Edit, to the downvoters. In order for me to produce better commentary on this topic in the future, I would like to know: Do you view these statements as incorrect, misleading, or simply unhelpful?
- yaantc 7y agoThe distributed operation I consider necessary for something to be called blockchain technology. Git is an example of a Merkle tree [1], which definitely is a key component of a blockchain but not enough on its own. One also need a distributed consensus mechanism. The distributed operation is both necessary, and the key challenge for blockchains. There is a heavy cost to secure distributed operations. My personal opinion is that most people don't care so much about distribution, they're fine with a few trusted/regulated actors using more efficient centralized schemes. And this is, in a nutshell, the big challenge of blockchains IMHO. And why some people are tempted to drop the challenging part, put a centralize scheme on top of a Merkle tree and call it a blockchain to ride the hype. I don't think this is correct, both technically and ethically. It's "merklewashing" ;) [1] https://en.wikipedia.org/wiki/Merkle_tree https://en.wikipedia.org/wiki/Merkle_tree
- jart 7y agoLet's just call it merkle paxos but it's much more fun to read the original paper.
- kd5bjo 7y agoThat makes sense. So, continuing with the git/source control analogy, what’s missing is a distributed mechanism of both publishing new forks and then deciding which of those forks “wins”. It sort of happens now in an ad-hoc way, if you include blogs and such with big fork announcements that sometimes result in an effective change of ownership, but that’s a slow, rare, unreliable process at the moment.
- cachestash 7y agoIs your ssh key not a proof of identity?
- kd5bjo 7y agoIt can be, but git neither validates nor stores it. All of that is done by optional access control layers on the computer where the git repository resides.