Hacker News Reader

Comment by scyclow | Hacker News Reader

I feel like this article, among others, kind of misses the point on why blockchains are useful. Decentralization, IMO, is incredibly overrated. What makes blockchains useful is that you can build highly interoperable applications that won't break due to changing interfaces. If AWS or the US government built a similar general purpose, open smart contract platform, I think we'd see similar levels of activity there. But they didn't, so we're using blockchains instead.

View on HN

Standards and decentralization are orthogonal to each other.

Standards are good for interoperability. Decentralization, depending on the flavor, either focuses on distributed hosting ensuring the system not needing to be run by some huge international company, or on not having a center of authority that mandates some particular standard or agenda.

Many fail in the later aim. Bitcoin is supposedly decentralized but in practice controlled by a very small group, for instance.

the smart contract space is doing pretty well in terms of standards. the "tokens" and "NFTs" you hear about in the media are implementations of specific standards.

That doesn't make much sense. There's APIs now, and API intermediaries line Zapier, IFTTT. AWS APIs are backwards compatible, as I'm sure is the case for whatever US government APIs there are.

So what's missing?

If I write some dip shit application and host it on my server, and you want to write an application that interops with mine, I can change my API (or the data itself) under your nose. You have no formal guarantees. Smart contracts make guarantees about what can change and what can't change.

Ethereum makes no such guarantee after the introduction of delegatecall, where it's possible to replace one smart contract implementation with another.

True. I'm curious what this sort of hit replacement would look like in practice. But, you can't add it in after the fact, so you at least have the option of writing contracts that give you the formal guarantee.

Mostly agree, but missing the point that there are non-decentralized blockchains that operate like an open standard seem to failed to gain any traction (or at least awareness) since 2019, to name some: Hyperledger Fabric, Corda R3.

What does a private blockchain accomplish that an immutable ledger does not?

Nothing really. It's just one kind of immutable ledger. If you have a use case where decentralization of the infrastructure is important, you don't care as much about network effects, and you know all your counterparties, then a private chain might be suitable.

When you identify the primary reason why corporations and governments never built built a centralized smart contract platform that's general purpose, open source, cross-border and uncensored, then you will know why decentralization is essential.

Don't get me wrong, the decentralization is important. I'd rather the network infrastructure of an expansive, public, general purpose computing network be decentralized. But to be useful in the first place you need important, interoperable applications running on it.

why don't you tell us?

The author understand the usefulness of blockchains. The point is that there are technologies which are components of the blockchain that are really all you need to implement a lot of use cases. For ex pub/priv key alone.

Reply on news.ycombinator.com