Hacker News Reader

Comment by shuntress | Hacker News Reader

Block chains can be useful. Git is the most obvious example.

Git is not what most people mean when they say blockchains (e.g. the decentralization of the blockchain and the use of the consensus algorithm is implied)

Kind of like how "Crypto" now means gamblinging on speculative assets rather than the study of cryptography?

Schneier calls that a "private blockchain"; he's critical about what he calls "public blockchain" which he defines as essentially the blockchain+mining+cryptocurrency component.

Git and secure scuttlebutt use merkle trees and chain blocks cryptographically, so yeah, they technically use a blockchain.

In my opinion main difference between "technically use a blockchain" and the term blockchain as used in the article and in common parlance is a way to automatically determine by distributed consensus which chain is the "real" blockchain. That's what proof-of-work or proof-of-stake do, and git and ssb don't provide a way to do that, so this article is not about them.

When I'm criticising blockchains I often use the term "distributed consensus blockchains" in an attempt to avoid confusion with Git and certificate transparency logs and other useful data structures.

The term blockchain didn't really exist before Bitcoin, and there isn't confusion about what people mean when they use it.

I'm pretty sure that if we researched the origins of the "git is a blockchain" meme we'd track it back to the community of crypto currency sceptics which I know you closely follow.

"A blockchain can be a useful data structure like the one git is using" isn't intended as a helpful clarification. It's an attempt to push for a change to the meaning of "blockchain" as originally (and currently) understood, to advance slight-of-hand arguments such as "blockchain was invented in the 70s, there is nothing really new here".

Is Git on a block chain?

Each commit is hashed as a block that includes the hash of the previous commit, so, yes.

Though, to be more accurate, you would say thay each "repository" is a block chain rather than say that Git is "on" a block chain.

It is a blockchain. It's different from Bitcoin, Ethereum, etc. in that it doesn't use a trustless consensus method. And that's literally the only situations where cryptocurrencies make sense, situations where you need to be able to operate with zero trust of the other party/parties. Turns out those situations are a lot rarer than crypto advocates claim, and even when such situations exist crypto is actually pretty limited in how much it can do/solve.

The underlying data structure for a blockchain and git is a merkle tree. So by some looser definitions, yes.

Reply on news.ycombinator.com