Hacker News new | ask | show | jobs
by manuelabeledo 1519 days ago
> None of the "research" and use cases linked need blockchains or are made better by blockchains.

You can read at a speed I could only wish. Kudos.

And of course they don't "need" them. We don't necessarily "need" the cloud, Docker, or web frameworks either, yet here we are. Yes, there could be a new, completely open, distributed database that let multiple users monitor and inspect all historical changes, while allowing some others to make modifications in it.

We could just use Cassandra and build the whole distributed identity and storage infrastructure based on it.
1 comments
dmitriid 1518 days ago
> You can read at a speed I could only wish. Kudos.

I read the sensational headlines. They are as boring and as old as ever: it's the same "use-cases" over and over again, none of them require blockchains, and in none of them blockchain provided anything of value.

> Yes, there could be a new, completely open, distributed database that let multiple users monitor and inspect all historical changes, while allowing some others to make modifications in it.

You have just described a regular database with an open dataset.

> We could just use Cassandra and build the whole distributed identity and storage infrastructure based on it.

Yes. Yes, we could. And that's what all of those "researches" and "use cases" will inevitably end up doing: scraping "blockchain" and replacing it with MySQL, Postgres, BigQuery, Cassandra, MongoDB, sqlite etc.

link
manuelabeledo 1518 days ago
> Yes. Yes, we could. And that's what all of those "researches" and "use cases" will inevitably end up doing: scraping "blockchain" and replacing it with MySQL, Postgres, BigQuery, Cassandra, MongoDB, sqlite etc.

I must admit that this right here was bait, because one of the storage backends used in enterprise blockchains is indeed Cassandra.

That’s part of the issue here. Blockchain technology is just something built on top of pre-existing tools, like Docker or Spring. My point is that if we were to scrap abstractions just because they could be done at a lower level already, why stop at blockchains?

link
maccard 1518 days ago
An abstraction has to offer value or solve a problem over the technology it's abstracting. Docker solves a myriad of problems that VMs and LXC have, and k8s solves a pile of problems that docker has. Blockchain's primary advantages are features that _every_ use case I've read does not want, and enterprise blockchains (from what I've seen at least) avoids those use cases by removing the features that differentiate blockchains from a normal DB.

> I must admit that this right here was bait, because one of the storage backends used in enterprise blockchains is indeed Cassandra.

That's really interesting. So is one party responsible for hosting maintaining and running that cassandra data store?

link
dmitriid 1518 days ago
> Blockchain technology is just something built on top of pre-existing tools, like Docker or Spring. My point is that if we were to scrap abstractions just because they could be done at a lower level already, why stop at blockchains

This sentence reads like it was written by GPT-3. It makes sense on the surface level, but makes no sense when you try and understand it.

link