Hacker News Reader

Comment by colordrops | Hacker News Reader

From the article, about the supposed expert who is skeptical:

> For all that she has learned, Radia is first to confess that she is by no means an expert in the technology — admitting that she cannot distinguish exactly what makes something blockchain technology

The article is right about one thing though - a lot of people are talking blockchains up without having any idea about what they are or how they work.

Conversely, a lot of people, like Radia in the article, are talking blockchains down, without understanding them.

View on HN

Ya know, Radia Perlman has decades of experience working on networking protocols, I'm willing to bet she understands blockchain as well as anybody. The statement "I cannot distinguish exactly what makes something blockchain technology" should probably translate as "currently the term 'blockchain' is thrown about so loosely its meaning is obscure."

I think it really pertains to using merkle trees. If you take a deep look into how bitcoin actually is implemented merkle trees are the base with which it is formed around.

This. 1000x. The beautiful thing is if you approach the conversation as one about a merkle tree built on a consensus algorithm, while avoiding the term 'blockchain', it's still possible to get folks excited about the tech.

> folks

The subset who have CS degrees maybe

People who "understand blockchains" don't agree on what exactly distinguishes something as blockchain technology, and people (and companies) are claiming all kinds of things as using "blockchain technology".

> People who "understand blockchains" don't agree on what exactly distinguishes something as blockchain technology,

Does this assertion come from a person who understands, or doesn't understand blockchain? :)

Because if you don't understand it, I would take your assertion with a grain of salt. And if you think you understand it, but you make this assertion, it means you disagree with many people that understand it... which hints that maybe you don't understand it after all ;)

Hint: the author is suggesting almost nobody else who claims they understand it, actually does. Whether that includes the author is almost besides the point. Take a good look around at the price of Bitcoin right now and ask yourself: is this because a whole lot of people just read the white paper, or is it just hype from people who have no idea what they are talking about. Draw your own conclusion.

Isn’t blockchain just a type of data structure?

It's obscure, honestly. But most cryptocurrency enthusiasts use the term to mean a blockchain (the data structure) together with a protocol for "consensus" (acceptance of a single instance of that data structure), commonly proof of work / "Nakamoto consensus".

It’s a linked list

Technically, it's just a linked list that uses hash pointers.

Nowadays, it appears to be used to refer to a distributed database that use blockchains or merkle trees (binary trees with hash pointers) and has a consensus algorithm.

Yeah sure but you need to understand asymmetric cryptographic to know how secure this data structure is, and how it can prevent attacks from private-key holders and non-private-key holders that claim to be holders.

No, “block chain” itself just refers to the hash-chain ordering of transactions. Using asymmetric crypto to verify spend conditions is sorta orthogonal to that.

blockchain without cryptography is like writing a data structure without any motivation for its existence. i.e. why hashtables were created in the first place? To have indexing (faster access than simple array/lists). What would be the point of a blockchain without cryptography? You cannot call that blockchain IMO (otherwise, any chain of blocks would be a blockchain, for example: a file is a blockchain, right?).

blockchain without cryptography is like writing a data structure without any motivation for its existence

Note that you said asymmetric cryptography and your parent said that asymmetric cryptography is not the basis of the block chain. And they would be right, at its core, the blockchain consists of blocks of transactions in a Merkle tree. The Merkle tree typically uses a cryptographically strong hash, otherwise it would be possible to modify the Merkle tree without anyone noticing. But a hash is not asymmetric cryptography.

In many applications of the blockchain, asymmetric cryptography is used to sign transactions. But that is largely orthogonal to the blockchain as a data structure. The blockchain as a data structure has useful applications without transaction signing, etc.

What makes the discussion clouded is that some people define 'block chain' as transaction blocks in a Merkle tree, while others also consider distributed consensus, etc. to be part of the block chain.

Also there's no reason asymmetric crypto has to be used. You could use Lamport signatures, for example.

I believe the comment you are responding is making a distinction between the public key cryptography used to authorize spending from a private wallet, and the act of including a cryptographic hash of the previous block in the current block creating a blockchain (which only requires a strong hash function, not asymmetric cryptography).

Blockchains as defined this way have existed long before there was a proof-of-work based consensus protocol to make Bitcoin a reality. In their more general form they are called "Merkel trees". Filesystems aren't generally structured this way but, for example, git repositories are Merkel trees.

Merkle. They were invented by Ralph Merkle, not the chancellor of Germany. This would be a Merkel tree:

http://c8.alamy.com/comp/D5AYH7/german-chancellor-angela-mer...

Git is an example of a blockchain without much cryptography, as someone else explained here.

The only cryptography is the SHA1 hash that links commits together, otherwise all cryptography is optional.

And that's totally okay because such repositories are usually run by people who have the authority over it.

A blockchain is a special kind of linked graph that uses cryptography to link each node, otherwise, there are no additional requirements and there are plenty of reasons you might not want anything else.

A lot of people know a lot about block-chains - but they know absolutely nothing about currency, finance or economics. I would say these people are scarier than IT managers who don't quite grasp crypto yet - because companies are not actually using crypto for much these days.

How is this comment not an ad hominem?

Many-to-most "blockchain experts" are hazy on where, exactly, the line between blockchain and not-blockchain falls (or, at least, there's enough ambiguity that drawing a sharp line is difficult to impossible).

To use that uncertainty to call out someone who professes skepticism about the general concept, while explicitly confessing that same uncertainty, as a means of disqualifying her perspective is therefore literally a fallacy of irrelevance.

how do we find an expert who balances both perspectives?

Reply on news.ycombinator.com